Pre-clinical evaluation of a replication-competent recombinant adenovirus serotype 4 vaccine expressing influenza H5 hemagglutinin

Abstract

Background: Influenza virus remains a significant health and social concern in part because of newly emerging strains, such as avian H5N1 virus. We have developed a prototype H5N1 vaccine using a recombinant, replication-competent Adenovirus serotype 4 (Ad4) vector, derived from the U.S. military Ad4 vaccine strain, to express the hemagglutinin (HA) gene from A/Vietnam/1194/2004 influenza virus (Ad4-H5-Vtn). Our hypothesis is that a mucosally-delivered replicating Ad4-H5-Vtn recombinant vector will be safe and induce protective immunity against H5N1 influenza virus infection and disease pathogenesis.

Methodology/principal findings: The Ad4-H5-Vtn vaccine was designed with a partial deletion of the E3 region of Ad4 to accommodate the influenza HA gene. Replication and growth kinetics of the vaccine virus in multiple human cell lines indicated that the vaccine virus is attenuated relative to the wild type virus. Expression of the HA transgene in infected cells was documented by flow cytometry, western blot analysis and induction of HA-specific antibody and cellular immune responses in mice. Of particular note, mice immunized intranasally with the Ad4-H5-Vtn vaccine were protected against lethal H5N1 reassortant viral challenge even in the presence of pre-existing immunity to the Ad4 wild type virus.

Conclusions/significance: Several non-clinical attributes of this vaccine including safety, induction of HA-specific humoral and cellular immunity, and efficacy were demonstrated using an animal model to support Phase 1 clinical trial evaluation of this new vaccine.

Figure 1. Ad4-H5-Vtn vector design.

The H5HA native coding sequence, with the polybasic domain removed (B), was derived from A/Vietnam/1194/2004 influenza virus and inserted into the Ad4 virus E3 gene region. The Ad4 virus E3 24.8K, E3 6.3K and E3 29.7K genes were deleted to accommodate the HA transgene and the splice acceptor site of E3 24.8K was retained to drive expression of the HA transgene. The E3A polyadenylation signal sequence, derived from Ad5, was placed downstream of the HA coding sequence. The use of a shuttle plasmid encoding the H5HA sequence and the Ad4 plasmid to obtain the final vaccine product is described in Materials and Methods.

Figure 2. Ad4-H5-Vtn virus infection of A549 cells induced H5 HA protein expression as detected by Western blot and flow cytometry analysis.

In the case of western blot analysis (A), A549 cells were infected in suspension with clarified crude lysate of Ad4-H5-Vtn virus from the 14^th passage. The cell extracts, 2.5% and 5%, were subjected to SDS-PAGE in 4–12% Bis-Tris gel, blotted onto nitrocellulose membrane and probed with anti-1194 H5HA polyclonal sheep antibody. Lysates as indicated, from uninfected A549 cells were used as negative control. Recombinant His tagged H5HA1203/2004 100 ng and 200 ng proteins were used as positive controls. β-actin was used as loading control. HA₀ refers to full length H5HA; HA₁ and HA₂ represent the proteolytically cleaved form of HA. In the case of flow cytometry (B), A549 cells were infected with a dose titration (vp/mL) of the Ad4-H5-Vtn vector: dotted line, 5×10⁶; solid line, 5×10⁷; dashed line, 5×10⁸. Infected cells were removed and subsequently incubated with primary anti-H5HA and secondary goat anti-IgG PE antibodies. A negative control, A549 cells infected with Ad4wt virus (5×10⁸ vp/mL) was included; grey fill.

Figure 3. Ad4-H5-Vtn virus growth is attenuated in various human cell lines versus Ad4wt virus.

Growth of Ad4-H5-Vtn virus was compared to growth of Ad4wt virus in several human cell lines: A549 a lung carcinoma cell line (A); Hu Tu 80 a duodenum adenocarcinoma cell line (B); MRC-5 an embryonic lung fibroblast cell line (C); H1299 a lung carcinoma line (D); and HepG2 a hepatocellular carcinoma cell line (E). Virus infection and cell growth conditions are described in Materials and Methods.

Figure 4. Vaccine-induced H5HA-specific humoral response in the presence and absence of pre-existing Ad4-specific immunity.

Mice were immunized i.n. with 1×10⁹ vp of Ad4wt virus per mouse to establish pre-existing immunity to the vector. Four weeks following the immunization, ten individual mice were bled and Ad4-specific neutralizing antibody titers were determined (A). Mice immunized with Ad4wt virus and naïve mice were subsequently immunized i.n. with a dose titration of the Ad4-H5-Vtn vaccine; 1×10⁹, 1×10⁸, 1×10⁷ and 1×10⁶ vp per mouse and bled 6 weeks after vaccine immunization and again 5 days later following H5N1 reassortant virus challenge to determine HAI antibody titers (B). The immune responses are represented by post-Ad4-H5-Vtn vaccine immunization (open bar) and post-H5N1 reassortant challenge (cumulative of open and black fill bar). Three mice from the group were bled and sera pooled to determine HAI antibody titers.

Figure 5. Vaccine-induced H5HA-specific cellular response in the presence and absence of pre-existing Ad4-specific immunity.

Mice were immunized i.n. with 1×10⁹ vp of Ad4wt virus per mouse to establish pre-existing immunity to the vector as previously stated. Two mice were sacrificed and splenocytes pooled to determine Ad4wt virus-specific cellular immunity, as assayed by IFN-γ ELISPOT (A). Two mice were also sacrificed 6 weeks after vaccine immunization and again 5 days later following H5N1 reassortant virus challenge and splenocytes pooled to determine H5 HA-specific cellular immunity evaluated by IFN-γ ELISPOT specific for four H5HA-derived 15-mer peptides, stimulating peptides for ELISPOT response (B). In the case of part (B), the immune responses are represented by post-Ad4-H5-Vtn vaccine immunization (open bar) and post-H5N1 reassortant challenge (cumulative of open and black fill bar). Mice pre-treated with Ad4wt virus and challenged with H5N1 reassortant virus demonstrated no detectable H5HA-specific cellular responses 5 days post-influenza virus challenge. An asterisk * denotes significant IFN-γ responses, p<0.05. In the case of two asterisks (*/*) associated with a bar, the bottom and top asterisks refer to post-immunization and post-reassortant H5N1 virus challenge, respectively.

Figure 6. Mice immunized with Ad4-H5-Vtn vaccine lost no weight, survived a lethal H5N1 reassortant viral challenge and presented with a reduction of H5N1 reassortant virus in the lungs.

Groups of mice were immunized with Ad4wt virus to establish pre-existing immunity as described previously in Materials and Methods section. Mice were subsequently immunized intranasally with a dose titration of the Ad4-H5-Vtn vaccine. Six weeks following Ad4-H5-Vtn vaccine immunization, the mice were challenged with a lethal dose of H5N1 reassortant virus. If animals were recorded to have lost 20% or more of their original weight for two days in a row they were euthanized. Weights of the mice were evaluated daily (A), and survival of mice were evaluated over a 14 day period (B).

Figure 7. Mice immunized with Ad4-H5-Vtn vaccine presented with a reduction of H5N1 reassortant virus in the lungs.

Groups of mice were immunized with Ad4wt virus to establish pre-existing immunity as previously described in Materials and Methods section. Mice were subsequently immunized intranasally with a dose titration of the Ad4-H5-Vtn vaccine. Six weeks following Ad4-H5-Vtn vaccine immunization, the mice were challenged with a lethal dose of H5N1 reassortant virus. Lungs were recovered from a subset of mice 5 days post-challenge to determine influenza-specific viral titers.